 Live from Santa Clara, in the heart of Silicon Valley, it's theCUBE, covering Juniper Nextwork 2016. Brought to you by Juniper. Now, here are your hosts, John Furrier and Stu Miniman. Hey, welcome back everyone, here live in Silicon Valley for a wrap up of the Juniper Network's annual User Conference Live coverage from SiliconANGLE Media theCUBE here in Silicon Valley. I'm John Furrier, Stu Miniman, and wrapping up here is Ethan Banks, co-founder of Packet Pushers. Moving packets, they're pushing packets, they're pushing boxes, they're all going software now. Welcome back to our wrap up. I wouldn't say they're all going software, but yeah. Well, Juniper's betting the ranch on NFD. Stu, great guest coming on here, the CEO came on, all the top dogs, some customers. This is Juniper's continued transformation, some new messaging, digital cohesion, makes sense, things are coming together, all the piece parts are available. Package that up, last year it was disaggregation, and then the day the network's where the action is, security's a problem, but this open sharing thing, a lot of stuff, what are your thoughts? Yeah, so John, first of all, I'm a little disappointed, I didn't have acronyms thrown at me all day. I mean, when I go to a networking show, I want to be talking underlays, overlays, MPLS, IPv6, if we really made that transformation. But no, to be serious, it's interesting that Juniper really up-leveled it from the start here. We're talking about digital cohesion and security. So we've been looking at security, security has gone from, it's always been an important topic, but security is like kind of job one in IT these days. Money is actually being thrown, big money is being thrown at it, and something that lots of companies need to make sure it happens. And I thought Juniper had a good story that put together. I'd love to, Ethan, I guess the first question I have for you is, as a networking guy, does this hit the right note for people in the networking industry, or is it a little bit too highfalutin? So first thing, he didn't say digital transformation during that keynote, he said digital cohesion, which, from a networking perspective, makes a lot of sense. Because what is it that is gluing together all of the information? It's the network, right? And so what is Juniper network's strength? It's that, it's building that infrastructure. And so the vision that he put together, sharing digital cohesion was about, let's get all of these disparate infrastructures wherever they are, whether it's our data centers locally, or whether it's up in the cloud. Let's make it very simple to put all of those data sources together, connect them together in an easy way. And now we've got a platform upon which we can share information. So digital cohesion, the way, the takeaway I had was, it's really an abler for that digital transformation. It's digital glue would be another term, but that's not as good as digital cohesion from a marketing standpoint. And it does package it up nicely. It gives that North Star vision kind of feel. It is a bit marketing-ish, but it hangs together. It's not over the top. And again, it's not played out either. I mean, digital transformation is so played. It's one of the marketing terms that I felt made real sense, especially coming from the Juniper Network's point of view. So I like that element of the message resonated with me. The security story is another interesting one, because it springboards off of that cohesion vision and says, well, what if we take all of these building blocks we have for security, and then we put them together in our enterprises, in our companies, and want to have a cohesive security strategy where it's not enough to have, as the legacy security infrastructure, firewall at the edge, that hard, difficult edge to penetrate, to get into that, what is now would be the chewy middle that's easy to get around in. We don't want that. We actually want to have a cohesive security structure where we've pushed security all the way out to the edge that that perimeter is now everywhere. That's a big deal. I want to get you guys' thoughts on this, because one of the things that jumps out, I mean, do we go to a lot of shows? We see all the conversations, different perspectives, obviously, it's the networking company, now the glue layer, cohesive, but they have to win that 20 year battle, the Cisco install base, and I like how they're positioning themselves and they're being disruptive to that pre-existing market of networking, but yet the future scenario of NFV and the scenarios I talk about with self-driving cars, all this new digital stuff, they have to enable. And the players that are doing that are the big cloud guys. And so the telcos are ripe for a product and services software to do that. And so I want to get your thoughts on the dynamics of that telco to compete with the other big cloud guys, because together they're one big cloud. The second thing I'd like to get your thoughts on is, is there a social network developing in the security industry at the networking and inside the stack where the data sharing is becoming, as Jonathan Davidson pointed out, I like his view, a club, there's an organic development going on where people are sharing and peering data almost, if you will, as a social network, just to get more data. So thoughts on the how to disrupt the Cisco's and this notion of social sharing. Yeah, it's interesting, because you've got this dynamic on the one hand, there's a lot of open source, there's a lot of things where peers always want to talk as to what's going on. I was at an event a couple of weeks ago in like Red Bull Racing, was like all these cool things we're doing, I can't tell you the order of magnitude of how many sensors I have, I can't tell you order of magnitude of data I have, because that's my differentiation, proprietary-ness. And even, we had one of the Juniper ambassadors on and he's like, oh, security's really important, but mom's the word on anything else there, because that's my business. So we still always have this dynamic of, sharing's good, hear a lot here about that, but there's certain pieces, what's important to my business? And I think one of the things I've been happy to see over the last couple of years is, IT really sorting out what is important for their business, what is the differentiation that drives what they're doing and what's the stuff that they're just going to push off to either a platform or a vendor, because spinning up email servers is not something that is for, other than a handful of companies, differentiation out there. Networking is the glue for all of this and there's pieces of networking that's important. And you think you like the software strategy that they have, you think that's a good strategy. Yeah, I do. Thoughts on that? So if I'm a business, why would I spin up my own email server anymore, right? It's way cheaper, more reliable, more consumable, readily, I can swipe that card and go with Google or whoever that provider is. What interest do I have in running an exchange server? I'm a guy that ran a lot of exchange servers back in the day. This is the value in that anymore. So there's a lot of blocking and tackling you do on-prem, you got to load all this stuff up and push that to the cloud. You're saying, just take that off the table. Why wouldn't you? I know. I mean, software services there, these are some obvious wins that businesses can take advantage of. In the networking space, one of the things that I've looked at is, when is it something that I need to worry about? When the whole network fabrics and SDN first came out, I was like, there should be a line in the sand that we can draw and say, if I'm above this line, I care about it from below this line, it's probably not something, I'm not deploying enough switches, I'm not having enough change in my environment that if I want some of those services, I should just get network as a service or some infrastructure provider, host or service provider, someplace else that can give me that. I'm curious, Ethan, yeah. Well, I mean, there's trade-offs you make here. So if you're a business that has been used to running your own infrastructure for a long time, you keep all your services and house, the thing that you had was control. And that's the thing that you see, that's what you give up, just as soon as you begin moving those services up to the cloud. So you lose control over that reliability factor, you lose control over performance, and now you're giving up all of that responsibility to your provider. And then just as soon as you need to have that control back, now it's hard, because you've got a black box that you're staring at, I can't, you don't have- You don't know if policies are on there, you know, if it's policy-based, whatever, automated. It goes to be in the cloud too, right? Sure, I mean, the security functionality, compliance, reporting, all of that stuff, you are very much relying on that cloud provider to, whether it's a software as a service or infrastructure as a service, to meet all of those requirements for you. And if they can't, what are you going to do about it? It's tough, you still are kind of locked into operations with those providers if you've committed to them and moved your data up. And it's not as simple as, I think that's the telco opportunity that Juniper has. They can go to the telcos and say, hey, you have on-prem stuff, you want to move stuff for economic reasons and other operational reasons or business model, CAPEX, OPEX, no problem. If you deploy the Juniper there, you can have the same automation on configuration and security in their clouds. That's kind of what I heard. Did I get that right? Is that kind of what they're saying? I mean, that's what I thought I heard. There is some of that, sure. So if it's infrastructure as a service and you want to stand up a virtual MX router up in Amazon Web Services, you can do that. And then you've got that operational control that's there. Different case to be made for just what that infrastructure is on the back end, if the latency meets your SLAs and so on, that's a different conversation. But yeah, you're getting it. Well, an important one, we have a source that I haven't reported on this yet, but they ran inter-cloud pings between Azure and Google and they all have direct connections with each other and the latencies are off the charts, different. Like really big, huge. So there really is no inter-cloud. You end up having to look at an internet, somebody like that that's got pipelines into all of these different clouds. And they got to be direct connections. And then you're relying on that cloud exchange service to give you that sort of performance and consistency that you need between clouds. Yeah. I mean, that's a big inter-cloud, but he said also, Rami said, hybrid cloud's not going away, one of his predictions. Absolutely right. Public cloud, super important. Hybrid cloud, absolutely NFV, big bet pushing forward, security will only grow. Ethan, automation was talked about a bunch. I think you did a podcast, maybe the data knots recently, talking about that. I mean, I know we've been talking about automation like my whole career in networking, so it's about time, right? Now, can we do it now? Yes. Now. So okay, so the big challenge here with automation and networking is that all the platforms on the back end are different. And so where we're at in networking as an industry in automation is, we're at a point where we're trying to get to some standardization of interfaces so that as we are automating, we have some predictable interface that we're communicating with. That's been the big hold up right now. You've got Cisco with one set of interfaces and Juniper with another and every other vendor with their own APIs and ways you can go about programmatically configuring that gear. So we have movements like OpenConfig, the OpenConfig group of customers that are putting together models and the ITF is involved in this too, modeling what the network and different aspects of network operation should look like so that we have a standardized, predictable way that we can now apply automation techniques to. So it's not a little different when I'm working on my Juniper infrastructure versus my Cisco versus my Arista. There's some homogenization happening there and that's a big win. That is really moving the ball forward. Wait, we still need standards. I thought open source got rid of all that. Yeah, I know, right? Standards are huge and what's happening is we're seeing the open source projects maybe leading the way but then taking the lessons they learn as they put code out and feeding it back into the ITF. So for example, OpenConfig, they're not off doing their own thing. They are taking their models and working with the ITF to standardize those models and contributing their opinions about what the models that the ITF are proposing should look like. Yeah, what about the other, if we talk standards, what about certifications? Juniper's doing certifications here. I saw a nice table of people grabbing books and really digging in on some of this stuff. How's Juniper doing on certifications and what's your latest on networking certifications? So certs are, they are the course of study that you will follow to make sure you learn the things that you wouldn't learn on your own. This is how I like to think about it. I've done lots of certifications over the years and it puts you in a position of, in my job, I know X, Y and Z but I never know A, B and Z because I don't need those for my job. My company doesn't use those services. Put me in a certification role. Now I have to know everything. A, B, C all the way through the alphabet right through to X, Y and Z. That makes you more knowledgeable as a network engineer. So Juniper's had for years a great certification program. They've got a ladder, they've got an excellent series of books behind them. They've got great testing, et cetera. It's one of the more mature certification programs and ladders that are out there and by ladder I mean you can start small and go as complex and detailed as you want right up through their expert programs. So yeah, there is certification testing here and I'm not surprised to see a bunch of people taking advantage of it. It's when you're very serious about learning a specific vendor product very well, you're going to dive into that, makes sense. I know one of the things I've been looking at for the last couple of years is, if you're traditional, if you're a company that's been around for a while, Juniper's been around for decades, how are you interacting with the public cloud? Juniper has a decent message there. They've got services that are fitting into the other environment. We've talked to them at Amazon re-invent, we see them at lots of these different shows. What's your take on how networking interacts with the public cloud these days kind of in general and anything specific that you've comments on for Juniper? So the trick is to make your public cloud experience seem like it's part of your private cloud. I want to poke at that, because our current experience today, for the most part, sucks. Don't we want to change it? Don't I want my private cloud environment to be more like the public cloud? Think about it from a workload perspective. I want to be able to take a workload and move it from wherever it is to wherever I want it to be without the network getting in my way. And by that I mean I need to have good transport, transparent connectivity, and I also need to have security profiles that follow me so that when it's in my data center and I know exactly how it's been secured that that security profile is going to follow it up into the public cloud as well. And so when you're looking at networking products that will take you into the public cloud, that's a lot of what you're looking for. Extensions from the environment you're familiar with to have that same warmth and familiarity in the public cloud. Yeah, so I guess, let me just say it. I understand I want to have policies that can be the same either way, especially security needs to be able to span view that what I guess what we've seen from a lot of people is that the operational experience of the public cloud is better today. And if I can have that kind of simplified experience in my own data center or my hosting environment, that might be easier because today. So when you're saying networking is a sucky experience, you're talking about just our operational experience in the data center? Rami gave the statistic we've been quoting for the last 15 years, which is 90% of our budget is spent on the keeping the lights on and we need to flip that. And we can't flip that through little incremental things. We need to change the way we do things completely. So that's what I'm poking at. But the challenges in the environments that we own and the data centers that we own where we have the responsibility for the care and feeding of the switches and the firewalls and all the rest of it that make up our data centers, we have been managing them in a same boring and tedious way for a very long time now and that operational process is ingrained in how IT does what it does. And it's very hard to change that. It's not that we don't have ways to change that because hyperscale and cloud guys have shown us exactly how these things are done. But networking for example. Not as easy as it looks. No, what? That's your point. It's definitely not as easy as it looks. You don't just download Kubernetes and all of a sudden start mapping all of your processes to go with that sort of an appointment model. You have to pick and choose your battle. It's a cultural thing too. I mean, we hear the analogy like stop treating pets, caring for pets instead. Think cattle. That's the metaphor, but that's hard. But it's not even hard simply from a cultural perspective. It's hard from an application perspective. If you want to treat applications like a cattle, that means you need to have an application that could be mapped to that sort of an infrastructure. And that's a challenge that a lot of enterprises have. They're running applications that were never designed for cloud scale operations. Even if they don't need the scale, they might need that operational functionality. But they didn't build the apps that way. They built them in a monolithic way. That means they're standing them up on a, they're used to scale up, not scale out. Juniper talked about self-driving networks and we might have some of the same challenges that we have to get to self-driving cars. Most of us, we're not ready. We understand that there's people out there that are drivers that we'd like to take off the road, just like I'm sure there's people running around data centers. We'd like to get them totally out of that job, but we're not ready to totally just say, oh, okay, automate all the things. Well Stu, that's a good point. To talk about what Ethan just said, it's interesting that the people are showing the way. So for instance, if you told me three years ago that there'd be self-driving cars, autonomous vehicles, I would have said, yeah, yes, sure. I'll give you 10 years. It's totally hype. Now Uber's showing some prototypes. Ford's going to ship in five years production, mainly around deploying and it's happening now. So mainly it's going faster. So people have to move faster otherwise. It's kind of an inevitable point of a forcing function. I think that is what I'm looking at here is that the same dynamics exist. Well, if I hold on to the old way, because the mappings are hard, yeah, keep my job or whatever it is, clutching on to that old way, someone else is going to just take it. I mean, that's kind of, I mean, is that a dynamic or is that my mixing and metaphors? I think there's some of that. I mean, I think there's some that just want to hold on to the way things are and they're resistant to change because they're resistant to change. But I think a lot more people, particularly in networking or coming around to the viewpoint of, wait a minute, this is just another iteration of what technology is always about. And technology is in a constant state of change. We are always moving to the next thing, whether it's as simple as an incremental upgrade or as complex as a massive project that's going to bring about a whole new system into our businesses. Tech is constantly on the move and changing and this move, where we're moving from managing individual devices to managing networks as a whole and typing things in one at a time on the command line to automating giant swaths of our network. That's just another change. It happens to be a harder one. And that's, I think, what's throwing a lot of people off. It's a scary one. Well, Rami's point is too soon. My takeaway from that was I broke into the business in the 80s and I remember that when I was young, what are those mainframe guys that's so dogmatic and they clutched onto the mainframe so long they would not change because and the operating budgets on the mainframes just on the support and maintenance IBM, I would say they're huge. And Deck had a big presence in the midis but they were kind of mainframes in my mind. What's huge, they wouldn't move to the client server and then internet working that became that new world. So is it the same dynamic? I don't know, it seems to be the same. They're dogmatic. I'm just the old way. There is some of that dogmatism. There's no doubt about it. At Packet Pushers, we've gotten some notes from people who say, guys, I hear what you're saying about busting silos and don't just live in the networking silo but talk to other people in other groups and get to know your developers and so on. But I have a job description that will not allow me to do anything more than what I am doing. And so they had to talk about a cultural barrier. Some people are even legally bound through contracts or federal law. SLAs, all kinds of compensation. They're not allowed to talk to anybody else. They can't. And so that's a cultural challenge that's very difficult to overcome. So sure, there is some of that dogmatism. Great stuff. Let's wrap it up. Let's go around the horn here. So let's get our take on Juniper's event here. Ethan, we'll start with you. Thoughts, observations, what's the vibe? What's coming out of this show? Let's take away. It's the second year for the event. It's very customer and partner focused. The information is going straight from Juniper management to the faithful. So you're getting very strong messaging about what Juniper is doing. Very strong security message this time around. And it comes across as quite believable. So you look at what Juniper's portfolio is and what their goals are, and it makes sense. And so you sit back as a customer, you sit back as a partner and go, yeah, I can work with that solution. And they're transparent, too. I can sell that. They open the books. They show you everything. Very much, yeah. Very, very transparent. Stu? Yeah, 34% growth in attendance year over year. I think pragmatic is a word that we kind of throw at an event like this, but with a strong message. I think last year we felt Rami really energized. The founders still involved. I've seen tweets of him floating around. I believe he's the CTO now, but good energy with the group. The customers, partners, all very engaged and energized. So really good to see. Would love to see the financials of trailing, of course, because year over year, Juniper's pretty much flat. But the company, we think we've got the direction. Rami says he's got the full support of the board as to where he wants to take the company. And we've seen messaging got lined up, products are all getting there and the proof should be in the pudding looking forward to watching it going forward. Yeah, my take in watching it is I think they're doing a great job of balancing, maintaining that existing business that they have, and again, while investing, but yet talking about a new way of doing networking and the software approach they're taking kind of goes back to Pradeep's early, over 10 years ago he saw this vision. I think NFV is the betting the ranch. That's my big takeaway. But the betting the ranch on NFV as that North Star and then going to get there with an investment. And the thing that comes again out here is I think they got the right sharing strategy. I heard Jonathan's comment about this club. I think there's going to be an ecosystem that's going to develop around security practitioners who want the network data, who will see value at the network layer and using that event data and other network data. So I think they're balancing the current, ship the core, speeds and feeds, get the performance throughputs out, but yet offering that future path. I think that's nice. Versus the we're groping for relevance and hoping that the world will spin in their direction. It's a nice balancing act and I think that management's got a good handle on that. So guys. Yeah, I just want to give Ethan the opportunity. You know, what's new with Packet Pushers? Where are you guys showing up these days? Where do people find out more? So packetpushers.net is home base for us. We run four different podcasts. We have the network break about news, the weekly show that's aimed at engineers, the priority queue, which is even nerdier if you want to dive a little deeper and then data knots where we cover cloud and data center and all things about infrastructure on that show. And packetpushers.net is where you go for that. We're working on a YouTube channel. Coming up, we're going to be doing some video and yeah, a lot going on, a lot of content. Video, and we interviewed Greg Farrow a couple of years ago. I said, yeah, video, you're like the Howard Stern of networking, the king of all media. John, we've got a lot coming up. All right. Ethan, great to have you. It's theCUBE. We've got a lot, we've got Reinvent coming up. We've got a lot of great events. IBM World of Watson coming up and a lot more this year. And again, over a hundred events for theCUBE. Stu, we continue the momentum. This is theCUBE signing off here in Silicon Valley at Juniper Networks, Networks 2016. I'm John Furrier, Stu Miniman. We'll be at the next event. Thanks for watching. I remember when I had such a fan.